The topic of the Intereconomics/CEPS conference for which this paper was written was framed as a question: convergence or divergence in the EU? I am prepared to give an unambiguous answer. That answer is yes.
Nominal versus real convergence
There has been nominal convergence in the EU since the advent of the euro in particular. Inflation rates have converged across Member States as a result of the decline of inflation in Southern European economies such as Greece and Italy. The standard deviation of annual inflation rates, expressed in percentage terms, fell from six percent in the early 1990s to less than one percent in 1999, where it has remained ever since (figures are for the 12 original euro area Member States).
Interest rate convergence was even more dramatic. Nominal rates on 10-year Greek government bonds fell from 18 percent (a full 10 percentage points above German levels) in the mid-1990s to just five percent (mere basis points above German levels) in 2001, when Greece adopted the euro, and to even lower levels and narrower spreads in the mid-2000s. Spreads then widened explosively following the bankruptcy of Lehman Brothers in 2008, with the development of banking problems in Ireland and elsewhere in the Eurozone, and following revelations in late 2009 of budgetary deception in Greece. Prior to that, however, the trend toward nominal convergence was overwhelmingly clear.
Nominal convergence occurred for reasons both good and bad. It was good in the sense that nominal convergence is a normal feature of an integrated monetary zone. A single monetary policy should make for similar risk-free interest rates across an integrated economic area. But nominal convergence was bad in that there was a tendency to confuse the elimination of exchange risk with the elimination of default risk. Ten-year government bonds are not always and everywhere risk-free. Yet this was the belief in Europe after the turn of the century. This misperception may have reflected confusion between exchange risk and default risk, where only the former had been eliminated by monetary union. It may have reflected the existence of zero risk weights and zero capital charges for banks holding sovereign bonds. It may have been confounded by the fact that the ECB, though applying different haircuts to different maturities of government bonds, reflecting their different degrees of liquidity risk, did not distinguish them by degree of default risk.
Have these problems been solved? Any confusion between exchange risk and default risk was dissolved by post-2008 events; this, at least, has been one beneficial effect of the crisis. In addition, in 2011 the ECB introduced a schedule of graduated valuation haircuts on assets rated BBB+ to BBB- , replacing the uniform haircut applied previously to these instruments. The new haircuts were at least as high as the old haircuts and in some cases higher. But regulatory capital charges for risky sovereign debt are still missing. Basel III foresees changing this, but governments, worried presumably about burdening weak banks, hesitate to implement such reforms.
In contrast to developments on the nominal side, there has been a lack of real convergence (convergence of per capita incomes) in the euro area in particular. For the 12 original members of the euro area, the coefficient of variation of per capita GDP in purchasing-power parity terms was essentially flat from the mid-1990s until the crisis; between 2008 and 2015 this coefficient of variation nearly doubled.
Several factors contributed to this post-2008 real divergence. First, there was the asymmetric impact of the China shock. China’s accession to the World Trade Organization and integration into the global trading system were good for Germany because it specialised in the production and export of machinery and transport equipment, for which China had a voracious appetite. But these same events were bad for countries such as Italy and Portugal, who had considerable product-specialisation overlap with China.
Second, Germany benefitted from the integration of Eastern Europe into the Single Market. German firms built supply chains through which they outsourced labour-intensive activities to Eastern Europe and specialised at home in skilled-labour intensive stages of the production process. Companies in other Member States, in Southern Europe in particular, failed to do likewise, whether because their products were less conducive to this kind of supply-chain management, because they lacked the same kind of historical links to the region, or simply because management was stuck in its ways.
Third, because of inefficient banking and financial systems, countries like Portugal that might have been expected to experience catch-up growth did not see productivity rise with capital inflows. Capital, instead of flowing into manufacturing, flowed into relatively inefficient, technologically unsophisticated service sectors.
Finally, as countries like Portugal, Spain and Greece were experiencing capital inflows, they also suffered from overvalued exchange rates. This phenomenon of capital flows from Northern to Southern Europe can be thought of as a corollary of the nominal convergence discussed above. This Dutch-disease problem made investment in manufacturing unattractive.
More precisely, investment was inadequate in both manufacturing and high-tech services. There was the failure of some European countries to invest more in the new generation of information technologies (IT) and to reorganise production to capitalise more fully on their productivity-enhancing aspects.
Have any of these problems been solved? There has been progress on most fronts. The China shock has diminished in the sense that China is no longer growing its economy and exports at double-digit rates. The production and export of manufactures has picked up in Spain and Portugal. Overvalued currencies are now less overvalued due to the weakness of the euro relative to the dollar, internal devaluation and the end of indiscriminate capital flows from Northern to Southern Europe. Banking systems have been strengthened and reformed in some – albeit not all – European countries.
The result has been some recovery of productivity growth across the euro area, specifically in crisis countries where such recovery was most sorely needed. Between 2014 and 2016, Ireland and Spain had two of the three fastest rates of total factor productivity (TFP) growth among the major Eurozone countries, according to AMECO. Portuguese TFP growth was respectable, at 0.6% per annum, while Greek TFP growth was 0.5%.
Lessons from Italy
This is not the case in another crisis country, Italy, however. Italian TFP growth in 2014-2016 was barely positive. How are we to understand the contrast?
The standard interpretation of Italian underperformance emphasises inflexible product and labour markets, in conjunction with an undercapitalised banking system saddled with nonperforming loans. The limitations of this interpretation are twofold. First, it fails to explain why the country’s product and labour markets are sclerotic. Second, it ignores the fact that the stagnation of TFP growth is not simply a legacy of the crisis and associated banking problems. In fact, Italian TFP stopped rising already in the mid-1990s.
The mid-1990s were, of course, when the United States experienced a productivity surge lasting about ten years and associated with the introduction of new information and communications technologies. Wholesale, retail and financial services were reorganised to capitalise on the productivity-enhancing opportunities afforded by these new technologies. The period saw the introduction of big-box stores, just-in-time inventory control and electronic internet-based banking, along with a host of less visible but equally revolutionary product and process innovations.
If we want to understand Italy’s lack of convergence, we must ask why it proved incapable of capitalising on these opportunities. The answer lies in the mismatch between the country’s inherited institutions and the requirements of new technology. By the mid-1990s, Italy had approached the technological frontier. From this point the economic challenge was to innovate, and in particular, to develop and apply new information systems. But the managers of family-owned and controlled firms, whether professionals or family members, were reluctant to put their patrimony at risk. They were reluctant to radically reorganise production in order to capitalise on IT. Outside investors could do little to pressure them, hostile takeovers and shareholder votes of no confidence being all but impossible in the Italian corporate governance system. State enterprises were under little pressure to reorganise. Banks, whether owned or influenced by municipal and regional governments, were reluctant to lend to unproven and radical new projects. Politicians benefiting from their cosy relationship with industry had no desire to disturb the status quo.
These institutional arrangements – family firms, state enterprises and government-controlled banks – served the economy well in the era of catch-up growth. Italy could grow rapidly in the third quarter of the 20th century not just because it emerged from World War II far behind the technological leader, with a per capita income only half that of the United States, but also because the institutions inherited and developed in that period were well-suited to the circumstances of the time. The country already had a handful of large firms, founded by prominent families, capable of emulating the capital-intensive, high-speed production methods pioneered by the United States. It had IRI, established in 1934-1935, enabling the state to supply those large industrial firms with cheap intermediate inputs. It soon had ENI to provide those industries with cheap energy. Credit for capacity expansion came from a financial system dominated by government-controlled banks. The political elite worked with industrialists and financiers to coordinate these moving parts. This system was suited to a period of catch-up, when the task was to channel resources into established sectors utilising established techniques.
Growth proceeded rapidly, as industrial firms emulated best practice as defined by the United States. While growth slowed after 1973 as a result of higher oil prices and a more troubled international environment, public debt rose, reflecting new social demands. Still, the debt problem was manageable so long as growth was maintained as it was through the mid-1990s; because import competition was limited, the lira could be devalued as necessary to restore competitiveness, and Italy remained some distance from the technological frontier. Doing familiar things in familiar ways still paid dividends, though less than before. It delivered sustained output and productivity growth, though slightly lower than the rate enjoyed in the third quarter of the 20th century.
By the mid-1990s, however, Italy approached the technological frontier where the challenge was to innovate and, in particular, to develop and apply new information systems. From this point on, the same institutional arrangements – family firms, state enterprises and government-controlled banks – that served the country well in the era of catch-up growth became obstacles to its continued expansion. Other elements of the institutional constellation, such as an education system better at training a handful of skilled engineers than at producing a highly literate and numerate labour force, a heavily indebted public sector with few spare resources to invest in R&D, and a service sector heavily cossetted by regulation further compounded the problem.
Participation in the Single Market and the euro were designed to shock Italy out of this unhappy equilibrium. By intensifying competitive pressure and removing the easy option of lira devaluation, they were intended to force firms to install professional managers, invest in information systems, reorganise and become more internationally competitive or else risk going out of business.
So why didn’t more firms respond? One answer is that history casts a long shadow. The institutional arrangements that had served the country well were now deeply entrenched. It was impossible to change some of them without also changing the others. You can’t strengthen corporate governance and shareholder rights without also reforming the financial system, for example. You can’t reform the financial system without also limiting the government’s involvement in the economy. Each of these prevailing arrangements thus poses an obstacle to changing the others.
From this flows the Italian electorate’s dissatisfaction with the political establishment. Rather than administering the shock therapy the economy requires, the political class, which benefits from the status quo, has incentives to defend it since its members share the rents accruing to incumbent companies. Even reform-minded leaders like Matteo Renzi were unable to crack this nut given the resistance of the political mainstream. Their failure and the chronic underperformance of the Italian economy ultimately caused the electorate to turn against the political establishment. Unfortunately, it is not clear that the beneficiaries of this political revolt – anti-establishment parties of the left and the right – have a clear sense of the problem or the will to devise a solution.
Crisis of institutions
My answer to the question of convergence or divergence is thus “both”. Since the mid-1990s and the advent of the single currency, the European Union and euro area have seen strong nominal convergence. The record of real convergence is chequered by comparison. Convergence of TFP and per capita GDP has varied with time, before and after the global financial crisis in particular. It has varied across countries, specifically in the contrast between Italy and its neighbours.
This contrast underscores the fact that the problem is not just a legacy of the global financial crisis. The failure of real convergence goes back further in history. It is fundamentally a crisis of institutions.
This does not mean that institutions have to converge across EU Member States for economic convergence to take place. It does mean, however, that the mismatch between the inheritance and the institutional imperatives of 21st century technology must be addressed. Different countries can address this problem in different ways. But if they fail to do so, real convergence will not take place.
* Text of remarks delivered to the Intereconomics/CEPS Conference “Economic Convergence or Divergence in the EU?”, Brussels, 9 October 2018.
- 1 The relevant figures are available in a number of other contributions to this special issue, so I do not repeat them here. For an especially good summary and presentation of the evidence see J.R. Franks, B.B. Barkbu, R. Blavy, W. Oman, H. Schoelermann: Economic Convergence in the Euro Area: Coming Together or Drifting Apart?, IMF Working Paper No. 18/10, 2018.
- 2 My focus is on Western Europe, the transition economies of Eastern Europe being a somewhat special case.
- 3 If one considers instead all 19 members of the present-day euro area, there was strong convergence up to 2008, reflecting catch-up growth in the Central European and Baltic members and also strong if unsustainable growth in Southern Europe, fueled by capital inflows, which halted in 2008.
- 4 O. Blanchard: Adjustment within the Euro: The Difficult Case of Portugal, in: Portuguese Economic Journal, Vol. 6, No. 1, 2007, pp. 1-21; R. Reis: The Portuguese Slump and Crash and the Euro Crisis, in: Brookings Papers on Economic Activity, Vol. 46, No. 1, 2013, pp. 143-193.
- 5 S. Calligaris, M. Del Gatto, F. Hassan, G. Ottaviano, F. Schivardi: Poor Productivity: an Italian Perspective, in: CentrePiece, Vol. 23, No. 1, 2018, pp. 6-9.
- 6 B. Eichengreen: The European Economy since 1945: Coordinated Capitalism and Beyond, Princeton 2008, Princeton University Press; B. Eichengreen: L’eredità del miracolo economico, Il Sole 24 Ore, 5 June 2018, available at https://www.ilsole24ore.com/art/notizie/2018-06-04/l-eredita-miracolo-economico-141913.shtml?uuid=AE62oYxE.
- 7 Istituto per la Ricostruzione Industriale (Institute for Industrial Reconstruction).
- 8 Ente Nazionale Idrocarburi (Italian energy company).